More than a quarter of patients with severe aplastic anemia remain pancytopenic despite immunosuppressive therapy, and others have suboptimal responses in at least one lineage. We have developed a program to ask whether use of a drug stimulating the c-mpl receptor, which binds the endogenous hormone thrombopoietin, could stimulate human hematopoietic stem cells in vivo. Murine, our own rhesus, and human data all suggest that thrombopoietin can stimulate primitive hematopoietic stem and progenitor cells to proliferate without terminal differentiation. The small molecule oral thrombopoietin mimetic eltrombopag was initially developed to overstimulate platelet production from marrow megakaryocytes to compensate for platelet destruction in immune thrombocytopenia purpura ITP). We originally developed a protocol utilizing eltrombopag in a phase 1/2 clinical trial for patients with refractory severe aplastic anemia. We reported that eltrombopag had efficacy in this setting with 44% (11/25) of patients having clinically significant hematologic responses (Olnes et al, NEJM 2012). We then reported additional safety and efficacy data on an expanded cohort of 43 patients, confirming an overall response rate or 40% with 3 to 4 months of treatment, including tri- and bilinear responses (Desmond et al, Blood, 2014). The majority of patients who remained on eltrombopag in an extension study (14/17) continued to show improvement, and 7 eventually had significant increases in neutrophil, red cell, and platelet lineages. Five patients with robust near-normalization of blood counts had drug discontinued at a median of 28.5 months after entry (range, 9-37 months), and all maintained stable counts a median of 13 months (range, 1-15 months) off eltrombopag. Eight patients, including 6 nonresponders and 2 responders, developed new cytogenetic abnormalities on eltrombopag, including 5 with chromosome 7 loss or partial deletion. This program resulted in FDA approval of a new labeled indication for eltrombopag in August 2014, and European commission approval in late 2015 for eltrombopag treatment of refractory aplastic anemia. This is the first new drug approved for aplastic anemia in decades and the first drug approved specifically for the refractory aplastic anemia patient population. During the current reporting period, we completed accrual and initial analysis of an additional study, asking whether more prolonged administration of eltrombopag to patients with refractory SAA, for 6 months instead of 3 months, and starting at a dose of 150mg/day instead of employing dose escalation, would improve response rate and rescue a larger fraction of refractory patients, based on the observation that patients in the initial trial improved markedly after the primary 3 month end point if they remained on drug. The response rate in 40 patients enrolled in this trial at the primary 6 month end point was 50%, not significantly improved over the initial trial. However, 25% of patients (5/20) did not reach response criteria at 3 months but were responders at 6 months, and thus were salvaged by the more prolonged treatment. Responses were faster than on the initial trial, due to starting at full dose, in contrast to dose escalation in the original trial, with no additional toxicity. We have combined both cohorts for molecular analyses asking whether EPAG impacts on clonal progression, and whether HPSC carrying mutations in specific genes linked to myeloid malignancies or other clones defined by new somatic mutations are specifically stimulated by EPAG treatment. Whole exome sequencing was performed in baseline samples and 6 month samples in all responding patients, and in any patient that developed cytogenetic abnormalities while on study. We found no evidence for specific clonal expansions related to EPAG treatment. The cytogenetic progression rate in both trials combined is now 19%. Of note, all patients that developed monosomy 7 or other deleterious chromosome 7 abnormalities did so within the first 3-6 months, suggesting potential preferential stimulation of a pre-existing clone. Other cytogenetic abnormalities appeared later, were often transient whether on or off drug, and were not accompanied by dysplastic changes or leukemic progression. During the reporting period, we also completed accrual of 34 patients to a clinical trial examine safety and efficacy of eltrombopag in patients with moderate aplastic anemia or unilineage cytopenias. The response rate was 50% at the primary endpoint of 4 months, including in a patient with Diamond-Blackfan Anemia and a patient with a telomeropathy. Patients tolerated doses up to 300mg/day without significant toxicity. We noted a lower rate of cytogenetic progression (6%) than in the prior trials in severe refractory aplastic anemia, and no patients with chromosome 7 cytogenetic progression. 65% of patients were able to have EPAG discontinued for a robust multi lineage response, however the majority needed to have the drug restarted to maintain counts, in contrast to our prior experience in refractory severe aplastic anemia. Based on data generated in all three trials, we made the observation that eltrombopag acts as a clinically-relevant iron chelator, raising blood levels of iron and resulting net iron loss in patients on the drug long term. In collaboration with Dr Andre Larochelle, we have elucidated the likely mechanism of EPAG activity in patients with high endogenous levels of TPO. This work will be described in his annual report. In collaboration with Dr Neal Young, we have reported on the kinetic relationship between TPO levels and response to EPAG in patients treated on multiple NIH bone marrow failure trials.